Image conversion apparatus and method

Abstract

Disclosed herein are an image compression apparatus and method. The image conversion method includes generating a multi-plane image reconfigured into layers in a depth direction based on multiple pieces of multi-view image data, generating an aggregated layer by aggregating the layers of the multi-plane image into at least one layer, and converting a multi-plane image including the aggregated layer into a two-dimensional (2D) atlas image.

What is claimed is: 
     
       1. An image conversion method comprising:
 generating a multi-plane image reconfigured into layers in a depth direction based on multiple pieces of multi-view image data; 
 generating an aggregated layer by aggregating the layers of the multi-plane image into at least one layer such that respective regions of pixels on each layer are packed into the at least one aggregated layer without overlapping between the respective regions of pixels; and 
 converting a multi-plane image including the aggregated layer into a two-dimensional (2D) atlas image. 
 
     
     
       2. An image conversion method comprising:
 generating a multi-plane image reconfigured into layers in a depth direction based on multiple pieces of multi-view image data; 
 generating an aggregated layer by aggregating the layers of the multi-plane image into at least one layer; and 
 converting a multi-plane image including the aggregated layer into a two-dimensional (2D) atlas image, 
 wherein the generating of the aggregated layer comprises: 
 generating a first aggregated layer from pixels that are seen first for respective identical pixel coordinates when an origin of a camera is viewed from a deepest layer among the layers; and 
 generating a second aggregated layer from pixels that are seen first for respective identical pixel coordinates, with exception of pixels included in the first aggregated layer, when the origin of the camera is viewed from a second-deepest layer among the layers. 
 
     
     
       3. The image conversion method of  claim 2 , wherein the generating of the aggregated layer is terminated when no pixel is present in each of the layers. 
     
     
       4. The image conversion method of  claim 2 , wherein the 2D atlas image includes a transparency image and a color image, which are generated based on pixel information of respective aggregated layers of the multi-plane image, and a layer index image having position information of the pixels on the layers. 
     
     
       5. The image conversion method of  claim 1 , further comprising:
 compressing the 2D atlas image. 
 
     
     
       6. The image conversion method of  claim 5 , wherein bitstream data is generated from the 2D atlas image using High Efficiency Video Coding (HEVC), H.263, or Versatile Video Coding (VVC). 
     
     
       7. The image conversion method of  claim 1 , wherein the multiple pieces of multi-view image data include image data having a 2D array. 
     
     
       8. The image conversion method of  claim 1 , wherein the multi-plane image is generated by generating planes for respective distances in a depth direction and rearranging corresponding pixel values and transparency values of pixels, for respective planes. 
     
     
       9. An image conversion apparatus comprising:
 a memory configured to store a control program for image conversion; and 
 a processor configured to execute the control program stored in the memory, 
 wherein the processor is configured to: 
 generate a multi-plane image reconfigured into layers in a depth direction based on multiple pieces of multi-view image data, 
 generate an aggregated layer by aggregating the layers of the multi-plane image into at least one layer such that respective regions of pixels on each layer are packed into the at least one aggregated layer without overlapping between the respective regions of pixels, and 
 convert a multi-plane image including the aggregated layer into a two-dimensional (2D) atlas image. 
 
     
     
       10. The image conversion apparatus of  claim 9 , wherein the processor is configured to generate a first aggregated layer from pixels that are seen first for respective identical pixel coordinates when an origin of a camera is viewed from a deepest layer among the layers and to generate a second aggregated layer from pixels that are seen first for respective identical pixel coordinates, with exception of pixels included in the first aggregated layer, when the origin of the camera is viewed from a second-deepest layer among the layers. 
     
     
       11. The image conversion apparatus of  claim 10 , wherein the processor is configured to terminate generation of the aggregated layer when no pixel is present in each of the layers. 
     
     
       12. The image conversion apparatus of  claim 10 , wherein the 2D atlas image includes a transparency image and a color image, which are generated based on pixel information of respective aggregated layers of the multi-plane image, and a layer index image having position information of the pixels on the layers. 
     
     
       13. The image conversion apparatus of  claim 9 , wherein the processor is configured to compress the 2D atlas image. 
     
     
       14. The image conversion apparatus of  claim 13 , wherein the processor is configured to generate bitstream data from the 2D atlas image using High Efficiency Video Coding (HEVC), H.263, or Versatile Video Coding (VVC). 
     
     
       15. The image conversion apparatus of  claim 9 , wherein the multiple pieces of multi-view image data include image data having a 2D array. 
     
     
       16. The image conversion apparatus of  claim 9 , wherein the processor is configured to generate the multi-plane image by generating planes for respective distances in a depth direction and rearranging corresponding pixel values and transparency values of pixels, for respective planes.